摘要
该文首先提出了一种使用GA、模糊推理及BP算法相结合训练前向多层神经网络的方法,并重点讨论了实现步骤,然后结合一个实例,给出了实验结果,实验结果表明了这一方法的最有效的特性是BP训练过程的快速收敛。
The method and procedure about applying combination of genetic algorithm, fuzzy inference and BP(Back Propagati6n)algorithm to train feedforward neural network are stressly studied first. Then, an example is given to show the most effective property of the method, that is, the fast convergence of BP training process.
出处
《计算机工程》
CAS
CSCD
北大核心
1997年第S1期252-255,共4页
Computer Engineering
关键词
遗传算法
模糊推理
BP算法
神经网络
Genetic algorithm
Fuzzy inference
BP(Back Propagation)algorithm
Neural network
